This R01 application seeks to elucidate the mechanisms by which antidepressant medications have limited efficacy in Late Life Depression (LLD) in order to develop new treatment interventions for this prevalent and disabling illness. We hypothesize that the presence of executive dysfunction (ED), which is common in depressed adults over 60, impairs the ability to form appropriate expectancies of improvement with antidepressant treatment. Greater expectancy has been shown to improve antidepressant treatment outcome and is hypothesized to be a primary mechanism of placebo effects. Moreover, white matter hyperintensities (WMH) on magnetic resonance imaging (MRI) are more prevalent in patients with LLD compared to healthy controls. It has been argued that WMH contribute to the pathogenesis of LLD with ED and decrease the efficacy of antidepressant medications by disrupting connections between prefrontal cortical (PFC) and subcortical structures. Vascular lesions to white matter tracts may also compromise the pathway by which expectancy-based placebo effects influence depressive symptoms. Expectancies reflect activation in PFC areas that may improve depressive symptoms by modulating the activity of subcortical regions subserving negative affective systems (i.e., amygdala) as well as those important in reward and hedonic capacity (nucleus accumbens and ventral striatum). Thus, LLD patients with ED and WMH may sustain a double-hit to their ability to experience placebo effects in antidepressant treatments: ED diminishes the ability to generate appropriate treatment expectancies, while WMH disrupt the physiologic pathways by which expectancies lead to improvement in depressive symptoms. To determine whether decreased antidepressant medication response in LLD patients with ED and WMH is caused by a loss of expectancy effects, we will evaluate 130 outpatients with LLD at baseline to determine their degree of ED (interference score on Stroop Color-Word Test), WMH burden (severity score on Fazekas modified Coffey Rating Scale derived from anatomical MRI), and white matter tract integrity (using diffusion tensor imaging [DTI]). Building on work from my K23 Award, we will manipulate participants' expectancy of improvement in an 8-week duration antidepressant trial by randomizing them between open administration of escitalopram (i.e., high expectancy) and placebo-controlled administration of escitalopram (i.e., low expectancy). The difference in antidepressant response observed between open and placebo-controlled medication treatment is a measure of the expectancy contribution to outcome, which is substantial in younger depressed adults but we hypothesize will be diminished in LLD patients with ED and WMH.